Hierarchical Behaviour Spaces
Michael Tryfan Matthews, Anssi Kanervisto, Jakob Foerster, Pierluca D’Oro, Scott Fujimoto, Mikael Henaff · 2026
Read Hierarchical Behaviour Spaces by Michael Tryfan Matthews, Anssi Kanervisto, Jakob Foerster, Pierluca D’Oro, Scott Fujimoto, Mikael Henaff (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-24558.txt.
Description
Matthews and five colleagues replace the single reward per option in hierarchical reinforcement learning with whole spaces of behaviours, letting a controller blend reward functions to shape richer policies. Tested in the NetHack Learning Environment, the method, called Hierarchical Behaviour Spaces, performs strongly. Follow-up tests trace its edge to wider exploration rather than deeper long-term planning, a result that cuts against common wisdom about why hierarchy helps.
Catalog record
| CREATOR | Michael Tryfan Matthews, Anssi Kanervisto, Jakob Foerster, Pierluca D’Oro, Scott Fujimoto, Mikael Henaff |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 27387 |
| LICENSE | http://creativecommons.org/licenses/by-sa/4.0/ |
| RIGHTS | CC BY SA |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.24558